Altered effective connectivity during performance of an information processing speed task in multiple sclerosis

Abstract

Background: Functional magnetic resonance imaging (fMRI) studies of persons with multiple sclerosis (MS) reveal distinct patterns of activation during task performance. We were interested in determining whether distinct patterns of effective connectivity would be revealed with Granger causality analysis (GCA).

Objective: To characterize directed neural connections in persons with MS during a processing speed task between brain regions known to be activated in healthy controls.

Methods: fMRI and GCA were used to examine effective connectivity underlying performance of a processing speed task in persons with MS. In total, 16 individuals with MS and 17 healthy controls (HC) performed a modified version of the Symbol Digit Modality Task (mSDMT) in the MRI scanner. Eight seed regions were selected on the basis of a priori data showing areas involved in mSDMT performance of HC.

Results: Behaviorally, the MS group attained a level of accuracy equivalent to the HC group, although they were significantly slower. While there was a great deal of overlap in the connections relied upon by both groups, the MS group showed significant differences in connectivity between critical brain regions. Specifically, the MS group had more connections from multiple regions to frontal cortices bilaterally relative to HCs.

Conclusions: Greater neural recruitment by the MS group relative to HC is consistent with the neural efficiency hypothesis, and lends further support to the notion that more connections must be recruited to maintain performance in the presence of brain pathology.